Areview of recommender system methods for drug response prediction

Cancer is a genetic disease that results when cellular changes and accumulation of different types of mutations cause the uncontrolled growth and division of cells. Since cancer is a disease of genetic complexity and diversity, the drug response for different patients can be different. The main reason for this occurrence is the difference in the molecular and genetic information of individuals, such as gene expression data, the type of mutation in the genome and copy number alteration information. These findings and achievements have recently made a significant challenge in the prediction of drug response for an individual patient in the research of precision medicine. Several recommender system-based models were proposed for predicting drug response. SRMF [1], CaDRReS [2] and DSPLMF [3] methods are comprised using seven criteria on two CCLE and GDSC datasets.